Children and young people in
gangs: a longitudinal analysis
Summary and policy implications
Juanjo Medina, Andreas Cebulla, Andy Ross, Jon Shute, and Judith
Aldridge
About this project
This project was a collaboration between the Centre for Criminology and Criminal Justice at
the University of Manchester and NatCen Social Research.
Research team
University of Manchester
Juanjo Medina, Senior Lecturer
Jon Shute, Lecturer
Judith Aldridge, Senior Lecturer
NatCen Social Research
Andreas Cebulla (now at Flinders University)
Andy Ross (now at the Institute for Social & Economic Research, University of Essex)
Vera Schneider (now at the World Economic Forum)
Wotjek Tomaszewski (now at the University of Queensland)
Stephen Webster, Head of Crime and Justice Research
About the Nuffield Foundation
The Nuffield Foundation is an endowed charitable trust that aims to improve social wellbeing in the widest sense. It funds research and innovation in education and social policy
and also works to build capacity in education, science and social science research. The
Nuffield Foundation has funded this project, but the views expressed are those of the
authors and not necessarily those of the Foundation. More information is available
at www.nuffieldfoundation.org
Contents
Introduction....................................................................................................................... 3
Aims of the study....................................................................................................
3
What is a gang?...................................................................................................... 3
Methodology...........................................................................................................
4
Findings and policy implications.................................................................................... 4
Recommendations for further research.........................................................................
10
References........................................................................................................................
12
Copyright © Nuffield Foundation 2013 | Registered charity 206601
2
Introduction
Youth gangs have received widespread political and media attention in the last decade.
Gang membership has been linked to violent crime among young people, particularly in
cases of gun and knife crime. More recently, this focus on gang membership among young
people has been amplified by the rioting in some UK cities in summer 2011.
Based on the suspected link between gangs and violent crime, government and different
agencies have begun to develop policy responses to gangs. In November 2011, the
government published its cross-government strategy to combat gang and youth violence.
Alongside this, police forces and local authorities are drawing on US models to tackle gangs,
including the formation of dedicated units and multi-agency teams.
This study aims to enhance the evidence base for the developing policy context. We have
used data from the Offending, Crime and Justice Survey (OCJS) to explore the
circumstances that lead to young people joining, remaining, and leaving gangs in England
and Wales. The OCJS provides the best nationally representative data available on young
people’s victimisation and offending in the UK.
Our findings, which are discussed in more detail in this briefing paper, are that gang
membership increases the chances of offending, antisocial behaviour, and drug use among
young people. This finding vindicates the current policy approach of treating gang
membership as a distinct part of crime prevention and youth policy.
However our findings also show the diversity within the different groups defined as ‘gangs’.
There are dangers in adopting an overly-general conception of ‘gangs’, namely the risk of
drawing young people unnecessarily into anti-gang policies (‘net-widening’), and the
widespread and counterproductive stigmatic labelling of youth. In light of this, preventative
and restorative interventions need to take care in differentiating between deviant youth group
types. Blanket interventions may have desired consequences in some groups but create or
exacerbate problems in others.
Aims of the study
Our study had three main aims:
1. to improve our understanding of the antecedents and consequences of gang
membership;
2. to enhance our knowledge about patterns of gang membership; and
3. to provide a more accurate characterisation of juvenile gangs in England and Wales.
What is a gang?
For our analysis we followed the operationalisation of gang membership based on the
Eurogang Network definition of street gangs as "any durable, street-oriented youth group
whose identity includes involvement in illegal activity”, as well as a variant of it that focused
3
on "durable, street oriented youth groups that engage in offending behaviour". 1 Respondents
were classified as gang members not on the basis of their own self-definition, but on the
basis of their responses to questions about the characteristics of their peer group (see
Weerman et al. 2009).
Methodology
We carried out both simple descriptive and more complex multivariate inferential analyses of
the Offending Crime and Justice Survey (2003-2006), a longitudinal survey of young people
in England and Wales. This survey followed what is often referred to as an accelerated
longitudinal (or ‘cohort sequential’) design, where a group of different age cohorts are
followed for a number of years. We focused most of our analyses on young people who were
10-16 in 2003 (the only age groups for which we had relevant gang membership measures
up to 2006) and analysed how their behaviour changed from that year to 2006.
It is important to highlight that household-based surveys by design exclude young people in
youth offending institutions and other locations. This means that by design this type of
survey excludes some young people at high risk of being gang members. It is, thus,
particularly important to compare our findings with those of studies that focus on high-risk or
criminal justice populations (see Bennett and Holloway, 2004; Wood, 2006).
Findings and policy implications
1. Gang membership increases the chances of offending, antisocial behaviour and
drug use, but it is not a sufficient condition for these outcomes.
Independently of pre-existing individual characteristics, gang membership increases the
chances of offending, antisocial behaviour, and drug use. This applies at the point of
joining a gang, but also extends to problem behaviour measured a year later. Our
findings, based on a propensity-score analysis comparing otherwise similar young
people, contradict the argument that gangs exert no independent effect on violence and
other offending (for details see Medina et al. 2013b). On the other hand, we did not find
evidence that joining a gang makes it much more likely that young people will be fearful
of crime, be the subject of violent victimisation, or suffer injury from violent victimisation
(as has sometimes being reported).
However, gang membership is not a sufficient condition for offending or other forms of
problem behaviour. Many young people in gangs do not report any offending. Previous
analysis of the OCJS reported the fact that 37% of gang member did not report a
criminal act in the year preceding the study (Sharp et al., 2006). We find similar figures.
Implications: Gangs are an important factor in shaping problem behaviour and should
continue to be acknowledged as part of crime prevention and youth policy and when
1
We slightly modified the EG operationalization so that young people that socialises with peer groups
involved in only illegal drug taking were not counted as gangs. See our discussion in point 7 of this
report for the rationale.
4
developing interventions and preventative programmes. Although policies that address
all the factors that shape youth offending and violence are essential (including welfare,
family and community strength and poverty), 2 it is important to recognise that schemes
that neglect gangs run the risk of ignoring an important factor in shaping problem
behaviour. However, it is important that these policies oriented to tackle gangs do not
equate gang membership with offending, something which the current tone of policy
discussions, as well as current practice, seem to do more often than not.
2. It is difficult to explain how gangs produce these outcomes, but it is likely various
mechanisms (i.e. change in routine activities and normative orientation) play a role.
Our propensity score analysis also suggests that young people joining a gang increased
the likelihood of spending time socialising in the street and of expressing support for prodelinquency values. Young people who joined gangs also became more committed to
their delinquent friends even if this would get them into trouble with their parents or the
police. Finally, joining a gang also leads to greater police attention, even adjusting for
other potential explanations for this police attention such as offending.
This resonates with American findings which suggest that gangs operate as a social
learning context in which negative values are transmitted and which favour a
disengagement from pro-social peers and commitments. But it also invites the possibility
that police attention may play a role in gang identity reinforcement (for details see
Medina et al., 2013b).
Implications: While we know gang membership affects problem behaviour, we know
less about how or why this happens. This makes it difficult to design interventions aimed
at affecting the group dynamics within gangs because we lack evidence on the relevant
process at work. These emerging and tentative findings may help to design better
interventions in the future. They also suggest that care needs to be exercised when
policing young people associated with known gang members. We discuss this latter
finding in greater detail further below.
3. Leaving a gang does not automatically and immediately leads to less problem
behaviour, but within a year there is a discernible effect. Additionally, most members
leave the gang in under one year.
This finding is consistent with previous British research that documents leaving the gang
and desistance from crime as gradual processes. For example previous ethnographic
work conducted by our research team (Aldridge and Medina, 2008; Medina et al., 2013c)
has suggested that leaving the gang is not equivalent with breaking all social ties with
gang members (which on occasions may render young people vulnerable to
"backsliding", particularly in the short term), and that those that aim to disengage from
the criminal activities of the group often face serious difficulties in getting rid of their gang
reputation among practitioners and community members.
2
We should not forget many of these factors have been linked by researchers to gang formation itself.
5
Moreover, although durability is key to what defines a gang, the membership of
individuals is fluid and unstable, with short-term membership (no longer than a year) the
most typical. The likelihood of membership increases to the age of 15 where it peaks,
and then quickly decreases. This is consistent with the pattern that emerges from U.S.
based research. There is, therefore, a considerable degree of “natural” disengagement
from gangs that should not be jeopardised by preventative or punitive interventions.
Implications: Policy-makers and practitioners should have appropriate expectations
about desistance when working with young people leaving gangs. This evidence should
be used to inform the definition and evaluation of programmes that aim to encourage
both the process of leaving the gang and desistance from crime. The key for intervention
work is to find ways to facilitate the process of desistance without jeopardising it when
there is evidence of backsliding in the short-term. Continued support and attention
should be offered throughout this process.
4. Problem and anti-social behaviour are the strongest predictors for joining a gang.
We also conducted analysis to identify risk factors for gang membership affiliation and
disengagement. We did this using a variety of multivariate modelling and machine
learning approaches (i.e., multilevel analysis, generalised boosted regression) that
control for various confounders. The strongest predictors of both joining and staying in a
gang were pre-existing problem and anti-social behaviour. Having a sibling who is also a
gang member and previous violent victimisation were significant factors for persistent
gang membership. There was a lack of evidence relating to parenting, but the parents
who knew their children’s peers (a proxy measure for parental supervision and
monitoring) were less likely to have children who joined and remained in a gang. These
risk factors are consistent with research carried out elsewhere in Europe and the US.
Neighbourhood characteristics, such as deprivation and young people’s perceptions of
the area, were not found to be significant predictors of gang membership. 3 This is likely a
methodological artefact. The OCJS, like many other youth offending surveys, was not
well suited to capture neighbourhood effects on behaviour. A different survey design is
required to understand the social ecology of gangs and community characteristics that
predict gang membership. 4
Implications: Our findings suggest problem behaviour should be the key focus of
identifying young people for preventative programmes. Other British scholars have made
similar calls on different grounds (Bullock and Tilley, 2008). Voluntary 5 family
interventions with young people perceived to be at risk of gang membership may want to
emphasise and strengthen the capacity of parents to supervise their children, although
3
This is in contrast with findings from the Edinburgh Study on Youth Transitions and Crime (see
McVie 2010) and with recent analysis of survey data from London (Alleyne and Wood, 2011). In this
sense, it is important to keep in mind the OCJS did not measure the single community characteristic
more likely to predict gang joining: the presence of gangs in the community. This may imply that
breaking up gangs will have longer-term community effects.
4
Please note that this study design cannot say anything about the factors that predict gang formation
(as opposed to gang joining). Most of the existing literature seems to suggest a greater impact of
ecological factors on gang formation than in gang joining (for a review see Klein and Maxson, 2006).
5
For an argument against coercive family intervention projects in this context see Shute, 2011.
6
we still need a better understanding of the relationships between factors operating in
different risk domains. We should not consider them in isolation and think that tackling
one will offer a magic bullet. Voluntary intervention models that treat the family as a
whole (such as functional family therapy or multi-systemic therapy) may be particularly
appropriate in this context. We should focus on early family-level intervention as a
means of addressing risk at both family and individual levels simultaneously and in ways
that may have beneficial ‘spill-over’ effects to all children in the family unit (for more
details on the potential of family interventions in this area see: Aldridge et al., 2009;
Medina et al., 2010; Shute, 2011).
5. Gang members who are more likely to offend are less likely to leave the gang.
Our multivariate analysis also suggest that gang leavers (as opposed to those who
stayed in the gang) exhibited less problem behaviour at an early stage; had more friends
who were not local; and had a lesser degree of perceived family financial stress. In the
same way that there is an element of self-selection in joining the gang, there is clearly an
element of self-selection in leaving the gang. In other words, those with greater
propensity to offend are less likely to disengage from the gang.
Implications: Programmes designed to facilitate the natural processes of
disengagement from gangs should consider the particular challenges posed by young
people with a more problematic behavioural background, particularly if their full network
of friends is bound by close territorial boundaries. Perhaps these young people may
benefit from access to extra-neighbourhood opportunities to travel and interact with
peers outside their own territories. 6 Again, these findings are consistent with a narrative
(see: Clarke, 2009; Churchill and Clarke, 2009) that emphasises a comprehensive
approach to family interventions and policy, aiming not only to strengthen particular
family dynamics such as parental supervision, but also to foster the financial capacity of
these families and their ability to cope with financial stress. It may also be more
disruptive of gangs to focus initially on those more likely to leave or who are easier to
disengage (cf. Kennedy, 2012).
6. There is no evidence that the number of young people joining gangs has increased
over time.
We found no evidence that gang members are becoming younger or that the prevalence
of joining gangs has changed over time in the seven age cohorts we examined.
Implications: Moral panics emerging over allegedly rising prevalence of gang and gang
crime is misplaced (see also Hallsworth and Young, 2008). This finding matches what
we know about trends in both youth offending from the Youth Justice Board surveys and
in victimisation from the Crime Survey for England and Wales. It also matches our
ethnographic observations in a North West city (Medina at al., 2013c).
6
Extra-curricular activities are generally only effective where the activity is both structured and
supervised. Also, some simplistic interpretations of these findings (e.g. suggesting wide catchment
areas for school) need to be aware of emerging British findings that identify the transition from primary
schools with small catchment areas to secondary schools with wide catchment areas as a problem
(see Bannister et al.,2010 and Aldridge et al., 2011).
7
7. There are diverse sub-groups within the groups commonly labelled as gangs.
Current survey measures (and policy uses) of the term gang, including our own
operationalization in the previous analysis, mask considerable diversity. Some groups
labelled as gangs only engage primarily in illegal drug taking, whereas others have a
stronger criminal orientation. We also found youth formations that may look like “gangs”
because they have leaders or territories or similar, but which did not engage in illegal
activities. We defined five types of youth formation, three of which have little or no
participation on illegal activities, and two which have some involvement in illegal activity. 7
No illegal activity:
• pro social groups (60%)
• pro social street groups (20%)
• quasi formal peer groups (4%)
Involvement in some illegal activity:
• drug user groups (11%)
• higher level offending groups (5%)
We argue that only the latter group deserves the gang label, 8 given its damaging
consequences. Similarly, contrasts with the Arrestee Survey suggest enormous variation
in the type of activities and structure between youth groups formations we may identify
as gangs through representative self-report surveys of children and juveniles and the
type of gangs we identify when we work with young adults and older adults that have
been arrested. Unsurprisingly criminal justice populations tend to exhibit a more serious
criminal record. Although this diversity may seem self-evident, it is a detail often lost in
the policing and policy framing of youth gangs. It would be troublesome to use the image
of gangs we obtain from criminal justice populations (less prevalent and more serious) as
representative of all gangs.
Implications: Much work remains to be done in describing and defining youth group
formations in the UK. There are dangers in adopting an overly-general conception of
‘gangs’, and doing this may lead to both ‘net-widening’ (drawing young people
unnecessarily into anti-gang policies) and widespread and counterproductive stigmatic
labelling of youth (see Bullock and Tilley, 2008). Preventative and restorative
interventions need to take care in differentiating between deviant youth group types:
blanket interventions may have desired consequences in some groups, but create or
exacerbate problems in others (for a discussion see Maxson and Klein, 2006).
7
We used a technique called latent class analysis for this purpose. This technique classifies
respondents in differentiated groups upon the basis of their responses to a predefined set of survey
questions. The questions we used included different group characteristics in terms of their structure,
behaviour, and normative orientation.
8
Not all gang researchers would agree with this assessment.
8
8. Our findings suggest that police practice tends to focus attention, such as stop
and search and stop and account, on two groups: individuals with prior police
contact and individuals whose friends have had prior police contact - the "usual
suspects" (McAra and McVie, 2005).
Our propensity score models mentioned above (see section 2. in particular) noted that
gang membership increases unwanted attention from the police, net of other factors
(including self-reported offending). This finding led us to carry out a series of multivariate
analysis, which although outside the scope of the original project, we report here and in
section 9. for their relevance in this context. This additional analysis was not funded by
the Nuffield Foundation (for more details see Medina et al. 2013a).
These analyses suggested that having delinquent friends and prior contact with the
police seem to be better predictors, more consistent across our models and generally
have a stronger effect, than one's own self-reported offending behaviour. In particular, it
seems to matter more and more consistently across our models who you hang out with
rather than what you do. There is indeed clear evidence of "policing by association"
when it comes to gangs (see as well Bullock and Tilley, 2008; and Ralphs et al. 2009). It
seems as if in practice, "intelligence-led" policing means that hanging out with the "wrong
crowd", including gangs, is likely to exacerbate the chances of a young person being
approached by the police regardless of their own level of offending. These findings also
connect with the long critical criminology tradition exploring self-fulfilling prophecies
about policing.
Implications: "Intelligence" products, such as the databases of gang members and
associates that have started to penetrate British policing practice (and are subject to little
quality or legal control), have the potential to magnify the disparate treatment of those
who have offended in the past, or who happen to be friends or family members of
individuals engaged in offending and known to the police. Our findings here suggest the
need to better regulate these intelligence practices and products (see a discussion of
suggested good practice in the American context: Barrows and Huff, 2009).
9. Ethnicity seems to matter more than self-reported offending behaviour for
explaining the probability of being stopped and searched.
Our multivariate analysis of unwanted police contact mentioned above is one of the few
British studies that assesses the probability of being stopped and searched, adjusting
simultaneously for likely confounders such as self-reported offending or drug use (as well
as other possible explanations such as presence in the street). Ethnicity still matters for
explaining the probability of stop and search. In fact, its effect size is considerably larger
than that of other variables, including offending. Controlling for other potential
explanations, the odds of being stopped are over four times higher for Black young
people than for White young people. Nonetheless, sample limitations shade the precision
of this estimate. To the degree that the disproportionality ratio on stop and searches has
increased since 2005 (Equality and Human Rights Commission, 2010), it is possible
these effects would be more striking today.
9
Implications: These findings reinforce what many suspected. The racial
disproportionality in the application of police powers to stop and question citizens cannot
be explained away by hypothetical racial differences on offending. Practices that result in
these disparities and that may foster attitudes of legal cynicism and distance from the
police (Fagan and Tyler, 2005) need to be supported by clear and strong evidence of
positive impacts for public safety.
Although many police practitioners argue that stop and search is an effective crime
control tool, there is no experimental evaluation of its effectiveness in England and
Wales (or elsewhere). Furthermore, the few non-experimental evaluations of public
initiatives that relied heavily on increasing stop and search, such as the Tackling Knives
and Serious Youth Violence Programme, offer less than conclusive evidence of their
benefits and actually raise some serious doubts as to what those benefits may be (Ward,
Nicholas, and Willoughby, 2011; see as well, Miller, Bland and Quinton, 2000). It is thus
imperative that the proposition - that these powers are beneficial for public safety - is
rigorously evaluated. To contend, as a recent Police Foundation (2012) briefing does,
that estimating the deterrent effect of stop and searches is simply too difficult is to
seriously underestimate the level of development of criminological research on targeted
deterrence and the capacity of social science to precisely evaluate initiatives like this.
Similar policing tactics have been evaluated elsewhere and their impact estimated using
randomised experiments. Nothing, other than a lack of political will, makes it difficult to
estimate these impacts. The recent policy movements to reconsider the use of these
police powers seem in line with this set of considerations.
Taken together we believe our findings suggest that early interventions to tackle gang
membership should be:
•
•
•
•
parsimonious, welfare-oriented and diversionary;
offered on the basis of demonstrable problem behaviour as opposed to real or
imagined social status such as gang membership;
focussed on identifying and assisting natural gang disengagement processes; and
alert to possible stigmatising or labelling mechanisms that may counterproductively
reinforce gang identities or exclusionary processes.
Recommendations for future research
This research project suggests a number of areas that require greater attention in the British
context:
•
We still have a very limited understanding in the UK about the neighbourhood
conditions that are linked to gang emergence and to joining a gang. Instruments like
the Offending, Crime and Justice Survey are not optimal for answering these
questions, so we need study designs that allow us to develop this understanding.
The Peterborough Adolescent and Young Adult Development Study (PADS+), the
Edinburgh Study of Youth Transitions and Crime, the Project on Human
Development in Chicago Neighbourhoods, and the MacArthur Foundation Studies on
10
Successful Development in High-Risk Settings (Denver and Chicago) provide leads
on how this can be achieved.
•
Our research examined data for a limited age-group and for a limited period. We
need longitudinal studies that allow us to assess change in gang membership and
offending over a longer period of time. Only then will we be in a position, for example,
to assess the long term impact of adolescent gang membership. There is emerging
evidence from the US that gang affiliation, even if brief, can cast a long shadow on a
number of developmental outcomes.
•
The core questions developed by the Eurogang Network that allow for
operationalisation of gang membership are brief and have now been adopted in a
number of countries and studies (including the International Self-Report Delinquency
Study-II). Their use in other government sponsored surveys, evaluation projects and
administrative data collection may provide useful information for expanding our
knowledge on this topic and to advanced evidence-led policy.
•
Although we have a clear sense that gang membership increases the risk of problem
behaviour, we do not have sufficient research on the particular group processes or
mechanisms that underpin these increases in the UK context. Few good prospective
British data exist on the group-level context of young people resisting, entering,
persisting or leaving gangs. The relative (and discriminative) importance of social
status, social learning and conformity processes, together with changing perceptions
of self- and group identity, deserve to be addressed, ideally in a mixed-methods
design.
•
Our work on disengagement from gangs needs to be complemented with studies that
assess more criminally-involved populations (such as those reflected in samples of
those arrested). It is particularly important to understand the factors that explain
disengagement from gangs and attempt more differentiation between young people
with different patterns or propensity to be at-risk of greater or lesser criminal
involvement. We also need a prospective study (ideally randomised control trials) to
show whether and to what extent natural (non-intervention mediated) disengagement
from gangs can be assisted (or hindered) by both criminal justice and or other model
intervention programmes. Furthermore, the distinction between desistance from the
gang and desistence from offending needs further attention. The Ministry of Justice
should seriously consider looking at the US Pathways to Desistance study as a
template for possible replication and extension in the British context.
•
Too little is known about how different youth groups may respond (differentially) to
policy interventions. More work is needed in this area. We also need to develop
research that assess the different developmental outcomes of affiliations with the
diverse groups we have uncovered, and that assesses how these groups may
change through time. In particular, we need greater scientific scrutiny of existing
police practice regarding gangs.
11
Unfortunately, the future of survey research on gangs (and more generally, youth offending)
is uncertain, at least in the UK. The OCJS was discontinued in 2006; and cognate Youth
Justice Board surveys also ceased in 2009. The remaining opportunity to ask gang-relevant
questions is in the Crime Survey for England and Wales, though this is with a relatively small
subsample (roughly around a 1000) of 10- to-15 year-olds, without a panel element, and
using a rigid definition that does not allow for international comparisons. What is worse is
that the two existing editions of the Crime Survey for England and Wales use two completely
different definitions of gang membership (with the more recent definition using looser more
ambiguous criteria with less grounding in the academic literature). And so, despite the UK
Coalition Government’s rhetorical emphasis on evidence-based gang policy (Shute,
Aldridge, Medina, 2012), we have, in fact, witnessed a major disinvestment in the production
of high-quality research data that constitutes a major part of this evidence base. We hope
the Home Office revisit its decision to support a national survey that allows the general
public and the scientific community to adequately monitor trends and development on youth
offending, whilst at the same time allowing rich theoretically-informed secondary analysis.
References
Aldridge, Judith and Juanjo Medina (2008). Youth Gangs in an English City: Social Exclusion, Drugs
and Violence: Full Research Report ESRC End of Award Report, RES-000-23-0615. Swindon: ESRC.
Aldridge, Judith, Jon Shute, Robert Ralphs and Juanjo Medina (2009). Blame the parents?
Challenges for Parent-Focused Programmes for Families of Gang-Involved Young People. Children
and Society. DOI: 10.1111/j.1099-0860.2009.00282.x
Aldridge, Judith, Ralphs, Robert, and Medina, Juanjo (2011). Collateral damage: territory and policing
in an English gang city. In Barry Goldson (ed.). Youth in crisis? Gangs, territoriality and violence.
London: Routledge.
Alleyne, Emma and Jane Wood (2011). Gang involvement: social and environmental factors. Crime
and Delinquency. DOI: 10.1177/0011128711398029
Bannister, Jon, Pickering, Jon, Batchelor, Susan, Burman, Michele, Kintrea, Keith and McVie, Susan
(2010). Troublesome youth groups, gangs and knife carrying. Edinburgh: Scottish Executive.
Barrows, Julie and C. Ronald Huff (2009) Gangs and public policy: constructing and deconstructing
gang databases. Criminology and Public Policy. 8(4): 675-703.
Bennett, Trevor and Katy Holloway (2004) Gang membership, drugs and crime in the UK. The British
Journal of Criminology 44(3):305-323.
Bullock, Karen and Nick Tilley (2008) Understanding and tackling gang violence. Crime Prevention
and Community Safety 10:36-47.
Churchill, H. and Clarke, K. (2009) Investing in parenting education: a critical review of policy and
provision in England. Social Policy and Society 9: 39-53.
Clarke, K.(2009). Early intervention and prevention: lessons from the Sure Start programme. In M.
Blyth and E. Solomon (eds). Prevention and youth crime: is early intervention working? Policy Press:
Bristol.
Fagan, J. and Tyler, T. 2005. Legal socialization of children and adolescents. Social Justice
Research, 18: 217-242.
Equality and Human Rights Commission. 2010. Stop and Think. A Critical Review of the Use of Stop
and Search Powers in England and Wales.
Hallsworth, Simon and Tara Young (2008). Gang talkers: a critique. Crime, media and culture. 4(2):
175-195.
Kennedy, David. 2012. Don’t shoot: One Man, A Street Fellowship, and the End of Violence in InnerCity America. New York: Bloomsbury.
12
Klein, Malcolm and Cheryl Maxson (2006). Street gangs patterns and policies. New York: Oxford
University Press.
Loader, I. 1996. Youth, policing and democracy. London: Macmillan Press.
Maher, Jenny (2009). Gangs? What gangs? Street based youth groups and gangs in South Wales.
Contemporary Wales, 22(1):178-195.
Medina, Juanjo, Robert Ralphs and Judith Aldridge (2010). Mentoring siblings of gang members: a
template for reaching families of gang members? Children and Society. DOI: 10.1111/j.10990860.2010.00307.x
Medina, Juanjo, Robert Ralphs, and Judith Aldridge (2013c). Gang transformations, Changes or
Demise:. In The Modern Gang Reader, ed. Maxson, C.L., Egley, A Jr., Miller, J and Klein, M.W, Los
Angeles: Oxford University Press, 2013.
McAra, L. and McVie, S. 2005. The Usual Suspects? Street Life, Young People, and the Police.
Criminology and Criminal Justice, 5/1: 5--36.
McVie, Sue (2010). Gang Membership and Knife Carrying: Findings from The Edinburgh Study of
Youth Transitions and Crime. Edinburgh: Scottish Executive.
Miller, J., Bland, N., and Quinton, P. 2000. The Impact of Stops and Searches on Crime and the
Community. Police Research Series 127. London: Home Office.
Police Foundation. 2012. Stop and Search. The Briefing, Series 2, Edition 3.
Sharp, Clare, Judith Aldridge and Juanjo Medina. 2006. Delinquent youth groups and offending
behaviour: findings from the 2004 Offending, Crime and Justice Survey. Home Office Online Reports
14/06.
Shute, Jon (2011). Family support as a gang reduction measure. Children and Society.
DOI:10.1111/j.1099-0860.2011.00368.x
Shute, John, Judith Aldridge, and Juanjo Medina. 2012. “Loading the policy blunderbuss.” Criminal
Justice Matters 87(1): 40-41.
Young, Tara and Simon Hallsworth (2011). Young people, gangs, and street based violence. In
Christine Barter and David Berridge (eds.). Children behaving badly: peer violence between children
and young people. London: Willey-Blackwell.
Ward, L., Nicholas, S., Willoughby, M. 2011. An assessment of the Tackling Knives and Serious
Youth Violence Action Programme. Research Report 53. London: Home Office.
Weerman, Frank, Maxson, Cheryl, Esbensen, Finn-Aage, Aldridge, Judith, Medina, Juanjo, and van
Gemert, Frank. 2009. Eurogang program manual. Eurogang Network.
(http://www.umsl.edu/ccj/eurogang/EurogangManual.pdf)
Wood, Jane. 2006. Gang activity in English prisons: the prisoners’ perspective. Psychology, crime
and law. 12(6): 605-617.
Publications linked to this project
Medina, J. 2013. Police-initiated contacts: young people, ethnicity, and the 'usual suspects'. Policing
and Society. DOI: 10.1080/10439463.2013.784301
Medina, J., Aldridge, J., Shute, J, Ross, A., Cebulla, A. 2013a. Measuring gang membership in
England and Wales: a latent class analysis with Eurogang survey questions. European Journal of
Criminology. DOI: 10.1177/1477370813475393
Medina, J., Cebulla, A., Aldridge, J, Shute, J., Ross, A. 2013b. Proximal adolescent outcomes of gang
membership in England and Wales. Journal of Research in Crime and Delinquency. DOI:
10.1177/0022427813496791
Shute, J., Medina, J., Ross, A., Aldridge, J., Cebulla, A. Risk factors for gang trajectories. Working
paper (available upon request).
13